ChatGPT Now Uploads Files Directly

OpenAI is rolling out a significant update that allows users to upload files directly into their ChatGPT conversations. This move enables direct interaction with documents, spreadsheets, and even images without leaving the chat interface. The feature is detailed on the OpenAI News platform.

Users can now analyze spreadsheets, edit existing documents, summarize PDFs, or process image data directly within ChatGPT. Supported formats include CSV, XLSX, PDF, DOCX, JPEG, PNG, and TXT files, among others.

How to Work with Files

To utilize this functionality, users simply start a chat and then open the tools menu to select "Add photos or files." Once a file is uploaded, users can prompt ChatGPT with specific tasks.

Examples include requesting summaries of reports, visualizing sales data, rewriting documents for clarity, or extracting key information into tables.

ChatGPT can also generate specific outputs like tables or charts, and users can download these generated files.

Connecting to External Tools

Further enhancing its utility, certain ChatGPT versions support app integrations. These apps allow ChatGPT to access third-party tools, bringing external data and context into conversations.

Users can find and connect these apps via Settings → Apps. After authentication, these apps will appear in the tools menu or can be invoked using @ or / commands.

For enterprise users, the availability of specific apps is managed by organizational administrators.

Data accessed through these apps is not used for training OpenAI models by default, ensuring business data privacy.